Will IBM ever convert Perspectives to use the Rest API?

[John Hammond]

Perspectives assumes a LAN where a small request can be immediately acted upon. The feedback from the response to your chat can be used to better direct your query. Citrix allows this behaviour to occur over a WAN by transmitting the display only, like the terminals in the 1980s. That said, it is a marvel of compression.

IBM has tried to replace perspectives with Performance Modeller which is still a thick client. I assume the logic was that PM would provide a superior interface, in that it would look modern. In actuality the interface was inferior and buggy.

Now we have PAW which is an improvement on PM. This is a web based interface that uses Javascript to give immediate feedback on a WAN. Some of the new functionality (eg hierarchies) is only available through PAW. Yet this interface is yet to gain the momentum to make it a 'well used' system, where the bugs have been driven out by overwhelming usage. The vast majority of the world still runs on the Perspectives/Citrix combo. It has the comfort of familiarity, and being super quick and rock solid. Also PAW and even PM never replicated the functionality of Perspectives.

In retrospect, I suppose, the PM path of revolution may not have been correct. Microsoft evolves all of its products with backward compatibility. IBM decided to make the quantum leap when maybe the best thing would be evolution.

Some of the things that cripple Perspectives could be easily dealt with such as switching off the entire local update when a process is run. Having failed with PM, would it still not be prudent to try and migrate the user base to a new paradigm, by simply moving perspectives forward into a web based product, given that while PAW is clearly superior to PM, it still may not gain the momentum to make users to make the jump from Perspectives?

[scrumthing]

Nope. That won't happen. Clear statement from IBM. Don't know what more I could add to that.

[Alan Kirk]

TLDR: Not a chance in Hades.

Perspectives assumes a LAN where a small request can be immediately acted upon. The feedback from the response to your chat can be used to better direct your query. Citrix allows this behaviour to occur over a WAN by transmitting the display only, like the terminals in the 1980s. That said, it is a marvel of compression.

True, but it's also pretty antiquated.

IBM has tried to replace perspectives with Performance Modeller which is still a thick client. I assume the logic was that PM would provide a superior interface, in that it would look modern. In actuality the interface was inferior and buggy.

And abandoned, thank the gods.

Now we have PAW which is an improvement on PM.

I still wouldn't blanking know, as per my recent discussion in another place (the Cafe) describes, despite the new "easier installation". Still, I don't give a flying fig about PAW. I doubt I'd use it even if I could get it to work, because from what I've seen of it Cubewise Arc is massively better in every material respect. The problem is that because IBM were too lazy to build a self-contained client (which Perspectives is), without PAW I can't get whatever the Excel addin is called this week to work.

This is a web based interface that uses Javascript to give immediate feedback on a WAN. Some of the new functionality (eg hierarchies) is only available through PAW. Yet this interface is yet to gain the momentum to make it a 'well used' system, where the bugs have been driven out by overwhelming usage.

Mmm. I don't think that the "low usage" argument can be used to explain the bugs. Arc usage is necessarily a subset of TM1 usage, and I can't recall the last time I came across a bug in Arc. Yet bugs in PAW seem to be a weekly feature. And some of them are just frapping ridiculous beyond the speaking of it.

Of course it is possible that Arc usage exceeds PAW usage by virtue of the fact that Arc "just works", whereas IMHO PAW is the software equivalent of sewerage plant on which a Lancaster has just dropped a Blockbuster bomb.

The vast majority of the world still runs on the Perspectives/Citrix combo. It has the comfort of familiarity,

True for existing users, but I think IBM is trying to look to the future. And IMHO failing spectacularly, since the future as seen by most users would be "minimum friction" between start and productive use. I may have my issues with Microsoft but they're getting a lot better at this these days, except for Windows Update. I may loathe Adobe, but they are excellent at this. PAW seems to go in the diametrically opposite direction.

and being super quick

Ahyyyyyyeeeehhhhh... in certain ideal circumstances. Even when I'm in the office I look at the speed of a large view returning in Perspectives (classic API) vs Arc (Rest API)... it's no contest.

Oh, and never, EVER open Subset Editor on a several thousand member dimension (even if you only have a subset of the elements selected) with the Properties window open, because you learn pretty quickly that that's a Really Bad Thing to do unless you're booked in for a 3 course lunch at the restaurant down the road.

and rock solid.

Pretty much, yes.

Also PAW and even PM never replicated the functionality of Perspectives.

And PM never will, since it's a walking corpse. The intention is to eventually have everything that Perspectives does in PAW. Some day. Over the rainbow. Mañana. It'll be coming Real Soon Now, in the software industry sense of the term. And it still won't ****ing install easily or properly.

In retrospect, I suppose, the PM path of revolution may not have been correct.

That's like saying "The charge of the Light Brigade may not in fact have been a tactically prudent exercise".

Microsoft evolves all of its products with backward compatibility. IBM decided to make the quantum leap when maybe the best thing would be evolution.

Yes and no. Evolution has its merits, but the change to the REST API and the attempt to open up the server were pretty revolutionary but a net positive (in the case of the new API, a massive net positive) and probably essential to TM1's longevity. As has been the case through a lot of TM1's life, IMHO, the client side then lets the team down. Again.

Some of the things that cripple Perspectives could be easily dealt with such as switching off the entire local update when a process is run. Having failed with PM, would it still not be prudent to try and migrate the user base to a new paradigm, by simply moving perspectives forward into a web based product, given that while PAW is clearly superior to PM, it still may not gain the momentum to make users to make the jump from Perspectives?

The problem is that Perspectives simply cannot handle hierarchies. It would need a ground up rebuild to be able to deal with that AND to transfer it over to using the RESTful API. What you would then end up with is an entirely new add-in whose only connection with the legacy Perspectives is a 1990's look and feel, but with all the potential bugs of a completely new product. IBM have said that there won't be any further work done on the Perspectives add-in; I'd say that's the main reason why.

[PavoGa]

That's like saying "The charge of the Light Brigade may not in fact have been a tactically prudent exercise".

I literally laughed out loud at this. And, in a coincidence, watched the movie with Trevor Howard and Sir John Geilgud this past weekend. Howard was over the top.

[Paul Segal]

I've not used PAW enough to form an opinion (like Alan I'm an Arc aficionado), but the bits of PAW that make it to PafE and PafE itself are being welcomed with open arms in my user community. Benefits include (the perception) of faster speed, only having to use the local install Excel rather than a Citrix version, and a generally up-to-date look and feel. I do have some niggles, but it's better than trying to persuade our IT organization to keep supporting a crufty Citrix install when they're trying to shut it down as we're the only ones using it.

[David Usherwood]

John Hammond wrote: ↑Wed Dec 09, 2020 10:01 pm
In retrospect, I suppose, the PM path of revolution may not have been correct.
That's like saying "The charge of the Light Brigade may not in fact have been a tactically prudent exercise".

Although my great-grandfather fought in the Crimea, and quotes in his journal the fudged up announcement by the commander of how well the Lights had done, I prefer the surrender radio broadcast made by Emperor Hirohito on 15 August 1945 - "the war situation has developed not necessarily to Japan's advantage". https://en.wikipedia.org/wiki/Hirohito#Surrender
But to return to the subject in hand. PAW for development is looking decent, but there are some nasty bugs atm. PAW for input and reporting is crippled because DBRW type reports (slices and active forms), which are far better than the alternatives as regards output quality and interactivity, STILL do not support named hierarchies - three and a half years after launch. My take is that IBM have found it too difficult and are trying to fudge a low level of dynamic drilling into Quick Reports. What are they thinking?

[John Hammond]

Alan you have to get a like button or something like that. Enjoyed your post.